This study used a neurological examination course integrated into the physical examination curriculum at a medical center affiliated with a university in Taipei. This study used a nonrandomized, open-label, active-control, parallel-group design with within-participant repeated measurements (pretest-posttest). Data were collected at the following two predefined time points: baseline before the course (T0) and immediately after course completion (T1), with motivation and technology acceptance measured once at baseline.

The study was conducted within the neurological examination curriculum for fifth-year medical students at a university-affiliated medical center in Taipei, Taiwan. Participant recruitment and data collection were conducted from August 29, 2022, to April 22, 2024.

Fifth-year medical students (clerkship 1) rotating through internal medicine and neurology were recruited during the study period. Because the assessments and questionnaires were embedded within the formal curriculum, participation was considered quasi-voluntary. The students participated in the regular basic clinical skills neurological examination course, with each class including approximately 8‐10 students. Because different clerkship blocks varied in size and instructional modality across the study period, the number of participants in each teaching group was not identical. Students were informed that questionnaire completion was optional and that declining participation would not affect course grades, clinical evaluations, or academic standing. Survey responses were anonymized and deidentified before analysis.

Inclusion criteria included fifth-year medical students enrolled in the neurological examination course during the study period. Exclusion criteria included inability to complete the pre- and posttests, failure to complete course procedures, or missing any major component of the study assessments.

This study was designed as a pragmatic educational evaluation embedded within the existing clerkship curriculum. Therefore, a formal a priori sample size calculation based on a prespecified effect size, α level, and statistical power was not performed. Participant recruitment and group assignment were determined by cohort availability and the preexisting teaching schedule during the study period, rather than by an optimal allocation strategy (eg, Neyman allocation). As a result, group sizes were unequal between the synchronous remote teaching and the in-person teaching groups.

Specifically, the control group consisted of students who received standard in-person teaching as part of the routine curriculum. The experimental group comprised students from the same academic year who were additionally invited to participate in synchronous remote teaching. Both groups received comparable instructional content and opportunities for practice and feedback, differing only in the mode of delivery. As participation in the remote group was based on additional invitation and voluntary agreement, group allocation reflects convenience sampling without randomization; therefore, selection bias and baseline imbalance cannot be fully excluded.

The experimental group used the Remote Clinical Presence System (RCPS), an enhanced telepresence setup developed for real-time remote clinical teaching, with permission granted for its use in this study. RCPS was designed to support synchronous remote teaching in clinical settings by preserving real-time interaction between instructors and learners [29]. Google Chrome was used as the web interface, and Google Meet was used for audiovisual communication.

RCPS enabled bidirectional, real-time audiovisual interaction, allowing students to observe neurological examination demonstrations and receive immediate feedback while performing examination tasks remotely. To facilitate transmission of fine motor actions and examination techniques—such as tendon localization, patient positioning, and reflex hammer handling—the camera setup provided a stable and close-up view of the examination field. Viewing angles could be adjusted during teaching sessions based on instructional needs, and instructors were able to request repositioning of the camera or repeat demonstrations to ensure adequate visualization of key procedural steps.

Before each teaching session, audiovisual connectivity was checked to ensure stable transmission. In the event of transient technical interruptions, instruction resumed after reconnection, and critical examination steps were repeated as needed to maintain instructional continuity. This instructional approach is consistent with prior reports on enhanced telepresence systems and synchronous digital platforms, which emphasize real-time interaction, demonstration, and feedback as essential elements for skill-based clinical education [29-32].

This study used a quasi-experimental design of “nonequivalent experimental group, control group with pre- and posttests” (nonrandomized controlled trial). The control group consisted of students who received in-person teaching in the classroom. These students observed the instructor’s interactions with standardized patients, demonstrations of neurological examination procedures, and the results. The instructor interacted with the students, observed how they performed the procedures, and provided feedback.

The experimental group participated in synchronous remote video learning using RCPS. These students also observed interactions, demonstrations, and results via video. The instructor interacted with them through video conferencing, observed how they performed the procedures, and provided feedback.

The neurological examination course was conducted over 2 weeks, with each session lasting about 60 minutes, using 2 standardized patient scenarios. One scenario covered the motor and sensory systems, while the other covered the cranial nerves and cerebellar systems. The goal was to train students to focus on appropriate examination strategies based on patient symptoms. The instructor demonstrated techniques, interacted with students, pointed out common errors, and allowed time for students to practice performing the neurological examination. The instructor monitored both in-person and remote students, offering real-time feedback.

The participants completed assessments electronically via a QR code–linked Google questionnaire at two time points (T0 baseline and T1 immediately after the course). At T0 (pretest), they completed background questionnaires, knowledge tests, self-reported confidence items, and baseline measures of motivation and technology acceptance using 5-point Likert scales based on the Motivated Strategies for Learning Questionnaire (MSLQ) [33] and the TAM [34]. At T1 (posttest), they repeated the knowledge and confidence assessments. In this study, self-reported confidence referred to their perceived comfort and readiness to perform the examination, rather than an objectively observed psychomotor skill. Self-reported confidence was measured using a 5-point Likert scale (1=not confident at all; 5=very confident), administered at both T0 and T1.

The course evaluation questionnaire and knowledge tests were reviewed by a panel of neurology educators to ensure content relevance, clarity, and alignment with course objectives. Revisions were made based on expert consensus before administration. The knowledge assessment was designed to cover key neurological domains and cognitive processes relevant to the course and was divided into pre- and posttests with comparable content and difficulty [35].

The participants also underwent an OSCE administered at both T0 (pretest) and T1 (posttest) using the same standardized checklist and scoring rubric. They conducted tendon reflex assessments on standardized patients at the following five specific sites: biceps brachii, brachioradialis, triceps brachii, knee jerk, and ankle jerk. The evaluation criteria for this test included the following components:

Neurologists, trained as examiners, observed the students and graded their performance using a standardized checklist and scoring rubric. Examiners were blinded to participants’ group allocation to minimize assessment bias and were not involved in the instructional sessions. Before the study, all examiners participated in a consensus meeting to standardize the interpretation of scoring criteria and ensure consistency in evaluation. Interrater reliability was assessed before study initiation, with 2 examiners independently rating 5 students; the resulting Spearman correlation coefficient was 0.70, indicating acceptable agreement. In this study, the only directly observed psychomotor skill assessment was the tendon reflex examination OSCE, which served as a focused proxy for selected neurological examination skills. This focused approach was chosen to ensure feasibility and consistency within the course context.

Motivation and technology acceptance were treated as baseline learner characteristics, whereas knowledge scores, self-reported confidence, and OSCE performance were analyzed as repeated-measures outcomes. Before inferential analyses, normality of continuous variables and their change scores (T1-T0) was assessed using the Shapiro-Wilk test, supplemented by visual inspection of histograms and Q-Q plots.

Because outcomes were measured at two time points (T0 and T1), within-group pre-post changes were examined using paired-sample t tests when normality assumptions were met; otherwise, Wilcoxon signed-rank tests were applied. Between-group differences in change scores were examined using independent-sample t tests or Mann-Whitney U tests, as appropriate based on distributional assumptions.

Motivation and technology acceptance were further examined using Pearson correlation analyses to assess their associations with improvements in knowledge and OSCE performance.

To address baseline imbalances inherent to the nonrandomized design, regression-based, covariate-adjusted analyses were conducted using an analysis of covariance (ANCOVA) framework, with posttest outcomes specified as dependent variables, teaching modality as the fixed factor, and baseline performance and prior neurology rotation experience included as covariates. To assess whether prior neurology rotation experience moderated the effect of teaching modality, a teaching modality × prior neurology rotation interaction term was additionally included in the ANCOVA models, and P values for interaction were reported. Adjusted estimates with 95% CIs were reported to aid interpretation of the magnitude and precision of observed effects.

Subgroup analyses by tendon reflex examination steps, anatomical sites, and prior neurology rotation experience, as well as correlation analyses, were conducted as exploratory, hypothesis-generating analyses. Given their exploratory nature, no formal correction for multiple comparisons (eg, Bonferroni adjustment) was applied; therefore, the potential risk of type I error inflation was acknowledged, and findings were interpreted cautiously with emphasis on effect sizes and CIs.

Given the low proportion of missing data, complete-case analysis was performed, and no imputation procedures were applied. Study data were organized using Microsoft Excel (Microsoft Corp) and analyzed with SPSS (version 22; IBM Inc). All statistical tests were 2-tailed, with a 95% CI, and statistical significance was set at P<.05.

This study was approved by the Taipei Medical University Joint Institutional Review Board (approval N202201046). Written informed consent was obtained from all participants before study participation. Participation in questionnaire completion was voluntary. Students were informed that declining participation or skipping questionnaire items would not affect course grades, clinical evaluations, or academic standing. Survey responses and study data were anonymized and deidentified before analysis. All participants received a convenience store gift voucher valued at 100 New Taiwan Dollars (approximately US $3.17) as compensation for study participation. No identifiable participant images or personal information were included in the paper.

This study evaluated the short-term effectiveness of synchronous remote teaching compared with traditional in-person teaching for neurological examination training among fifth-year medical students. Both teaching modalities were associated with improvements in neurological examination knowledge, tendon reflex OSCE performance, and self-reported confidence. Although the synchronous remote teaching group demonstrated modestly higher postcourse knowledge scores, no statistically significant between-group differences were observed in OSCE improvement after covariate adjustment. Exploratory analyses further suggested that motivational beliefs related to SRL may be associated with greater psychomotor skill improvement in synchronous remote learning environments.

The absence of meaningful between-group differences in tendon reflex OSCE improvement suggests that synchronous remote teaching may support short-term acquisition of selected psychomotor neurological examination skills under structured instructional conditions. One possible explanation is that tendon reflex assessment relies on standardized procedural sequences, visual observation, and repeated guided practice, all of which can be effectively demonstrated through synchronous video-based interaction [10,36]. In addition, the RCPS system enabled close-up visualization of tendon localization, examination technique, and patient positioning, while instructors provided immediate corrective feedback during skill performance. These findings are consistent with prior literature suggesting that synchronous remote teaching environments may support psychomotor clinical skills when real-time interaction and structured feedback are incorporated [17,36,37]. However, unlike many previous studies that relied primarily on self-reported outcomes or asynchronous formats, this study evaluated performance using an OSCE-based assessment focused on a specific neurological examination task.

As this study used a nonrandomized design, baseline imbalance in prior neurology rotation experience required careful interpretation. Participants with prior rotations demonstrated higher baseline self-reported confidence and greater OSCE performance, suggesting that prior clinical exposure may contribute to foundational procedural skills [38]. However, covariate-adjusted analyses and formal interaction testing did not demonstrate evidence that prior neurology rotation experience modified the effect of teaching modality on either knowledge or OSCE outcomes. These findings suggest that the observed learning outcomes associated with synchronous remote teaching appeared consistent across different baseline experience levels, although residual confounding cannot be fully excluded.

The exploratory association between motivational beliefs and OSCE improvement may be interpreted through the lens of SRL theory. In particular, self-efficacy and motivational regulation may support learners’ engagement, persistence, and responsiveness to feedback during psychomotor skill acquisition [39-41]. These associations appeared more pronounced in the synchronous remote teaching group, suggesting that remote interactive environments may place greater emphasis on learners’ active self-monitoring and self-regulatory capacity [38,42]. However, because these analyses were exploratory and this study was not designed to directly evaluate SRL mechanisms, these findings should be interpreted cautiously and regarded as hypothesis-generating.

In contrast, technology acceptance was not significantly associated with knowledge or OSCE outcomes. One possible interpretation is that, within highly interactive and instructor-guided synchronous environments, technology acceptance may function primarily as a prerequisite for participation rather than a direct determinant of learning effectiveness [21,43]. This distinction may help contextualize the role of TAM in synchronous psychomotor clinical skills training, although the present findings remain exploratory and do not establish theoretical mechanisms.

Collectively, these findings suggest that the effectiveness of synchronous remote psychomotor skills training may depend less on technology acceptance itself and more on how learners actively engage with feedback and regulate their own learning processes. From a theoretical perspective, this observation is broadly consistent with SRL models, which emphasize self-monitoring, self-efficacy, and strategic regulation as important determinants of learning performance in complex educational environments [22,23,42]. In contrast, TAM may be more useful for explaining learners’ willingness to adopt and engage with educational technologies than for directly predicting skill acquisition outcomes once active participation has been established [19-21]. Although causal mechanisms cannot be inferred from this study, these findings contribute to the growing literature exploring how learner characteristics interact with technology-mediated clinical skills education and may help inform future theory-driven investigations of remote psychomotor skills training.

From a practical perspective, synchronous remote teaching may provide a flexible option for delivering neurological examination training when face-to-face instruction is constrained by geographical distance, scheduling limitations, infection control requirements, or limited access to specialist instructors. Enhanced telepresence systems and synchronous video-based educational environments may facilitate real-time demonstration, observation, and feedback while expanding access to clinical skills education across distributed training sites [17,29,36]. Although further evidence is required regarding scalability and long-term effectiveness, the present findings suggest that structured synchronous remote teaching may serve as a useful supplementary instructional approach for medical schools seeking to maintain interactive clinical skills training while addressing logistical challenges in contemporary medical education [36,37].

Several limitations should be considered when interpreting the present findings. First, this study used a quasi-experimental design without randomization, which introduces potential selection bias and residual confounding. In particular, baseline imbalance in prior neurology rotation experience may have influenced learning outcomes despite the use of covariate-adjusted analyses and formal interaction testing. Therefore, the findings should be interpreted as associative rather than strictly causal. In addition, participants’ awareness of being studied may have introduced a Hawthorne effect [44], potentially increasing motivation and engagement across both groups. Several subgroup and exploratory analyses were conducted without formal correction for multiple testing, which increases the potential risk of type I error; accordingly, these exploratory findings should be regarded as hypothesis-generating rather than confirmatory. Furthermore, self-reported confidence represents a subjective learner perception and may not fully correspond to objectively measured psychomotor performance.

Second, the study evaluated only immediate posttest outcomes and, therefore, does not establish long-term retention or durable skill acquisition. In addition, the directly observed psychomotor assessment focused specifically on tendon reflex examination skills, which represent only one component of the neurological examination; therefore, the findings should not be generalized to broader neurological examination competencies without further study. The pragmatic design also did not include a formal a priori sample size calculation, and unequal group sizes resulted from cohort availability and scheduling constraints, which may have limited statistical power to detect small between-group differences. Finally, the synchronous remote teaching sessions were conducted in relatively small groups with substantial instructor interaction and individualized feedback; therefore, the scalability and effectiveness of this instructional approach in larger educational settings remain uncertain.

Under the conditions of this quasi-experimental study, synchronous remote teaching was associated with short-term improvements in tendon reflex examination skills, neurological examination–related knowledge, and self-reported confidence among fifth-year medical students. Although modestly higher postcourse knowledge scores were observed in the synchronous remote teaching group, no statistically significant between-group differences were identified in OSCE-assessed psychomotor performance after covariate adjustment. These findings suggest that structured synchronous remote teaching may support short-term acquisition of selected neurological examination skills when real-time interaction, guided practice, and immediate feedback are incorporated into instructional design.

To our knowledge, this study is among the first to evaluate synchronous remote teaching for tendon reflex examination training using OSCE-based psychomotor assessment and theory-informed interpretation grounded in SRL and technology acceptance frameworks. These findings contribute to the growing literature on synchronous digital clinical skills education by extending prior research that has primarily focused on asynchronous learning or subjective learner perceptions. However, because this study assessed only immediate outcomes and focused on tendon reflex examination skills, the findings should not be interpreted as evidence of long-term skill retention or generalized effectiveness across the full spectrum of neurological examination competencies. Future research should incorporate randomized designs, larger multicenter cohorts, and longitudinal follow-up to evaluate scalability, long-term retention, transferability to other neurological examination domains, and strategies to support SRL in remote clinical skills education.